To encourage users to fill in profile information by sending them CTA messages and request from peers through the notification center.
Time
Current Status
My Role
Project Type
β
Internship Project
Add new features
8 Weeks, July 2019
Online
Product Design Intern
Project Owner
We enable employees in the organization to search for other employees base on skills, interests, knowledge, and experiences.
.png)
To do that, I did some research to understand why users don't fill in profile information currently and drew these two insights.
The solution is the notification center and the messages we send to users through it. Sift will ask users for profile information and enables the users to ask each other for profile information.
β
.png)
Data was collected after its about 3 months online. The growth was surprisingly high.
.png)
It pops up and ask for user information once in a while. Problems:
Only show up on the homepage, where our users barely stay.
Far from the main operation area
Looks like chatbot that provides optional help information.
π10% users responded to the old CTA.
%20(1).png)
It opens automatically and show the call-to-action message if users didn't fill in all required information or there's a new request from other people. Advantages:
Users can find it on any page.
It's closer to the main operation area and thus easier to be noticed.
π48% users responded to the new CTA!
%20(1).png)
Users receive a direct ask to fill in profile. They were thinking:
"Does other use Sift?"
"Does it worth my time?"
We enable peers to ask for one user's profile information, and the user will be notified though emails and in-app messages.
Users who receives the message will know that:
People do look at my profiles.
Profile information is for better collaboration.
π45% users responded to peers' requestοΌ
.png)
I worked on this project as the project owner, and reported biweekly to 2 remote PMs (our CEO and Design Lead). As my internship ended, I hand over my design to a PM and he shipped it. Though I was the only one working closely on this project, I very often reached to other designers, PMs, Developers and Client Success and Marketing team for help and feedback.
Main areas of responsibility:
- Management (sprint planning, weekly reviews)
- UX research (data analysis, interviews and user testing)
- Exploring and validating ideas (brainstorming, collect feedback from stakeholders, prioritization)
- Design (notification center on the website, emails and message content)
a.They only invest in popular tools, and they don't know if sift worth.
b. They don't see profile's value to themselves, to team members.
c. They see other empty profiles and thought that was ok.
a. They don't know the expected content and tone of the answer.
b. They themselves don't know the answer to some questions, such as "What I do".
c. Some information can be hard to recall.
a. The tool we use is only accessible on the homepage, where users don't stay.
b. It's far from the main operation area and users don't see it.
c. Looks like chatbot that provides optional help information.
β³οΈ
Goal
To understand the current profile filling condition, and find out the influencing factors.
πΎ
Methods
I asked for the raw data from engineers and analyzed the data of 24k users on Numbers by myself.
πΊ
Process
I firstly defined and calculated some key user status indicators, such as "active user", "healthy company", and four levels of user profile completion state. Then I Β compared the data of users in many dimensions, for example, users from different companies, of different career length in the company.
Influencing factors of profile completion rate (company name and data removed)
π‘
Insights
Factors influencing users' profile completion condition:
1. User activeness;
2. User's companies' size;
3. The way we ask users to fill in profile. (They tend to react when we emphasize the value to employees instead of the company.)
β³οΈ
Goal
Answer the questions raised in the data analysis
βοΈ
method
I conducted interviews with 5 users who come from companies of different scales, with different levels of profile completion and activeness.
π§
Questions Asked
User scenarios dealing with Sift and profile; their thoughts and perceptions in the process; thoughts to the previous measures we had; and why they fill in or don't fill in different sections of their profile.
I brainstormed a dozen of ideas, iterated them according to the team's feedback, and prioritized them. It turned out that the notification center can be the best choice to go with at this time.
The brainstormed ideas
Paper prototypes are easier for team discussion
The team believed that those ideas might all be effective, and they all need to be implemented. However, we wanted to build them piece by piece so we needed to decide the roadmap. Before the prioritization, I came up with the metrics that we can use in this specific project.
Metrics of Evaluation and Prioritization
.png)
Three User Groups
Possible information to send to users in different stages
π‘
Prioritization metrics
1. The cost of building the message supporting system;
2. The expected effectiveness of the message.
π
The message to Build
Request from colleagues to fill in profile information
β³οΈ
Design Goals
Users will take actions to fill profile after reading the email.
π€¨
User concerns learned from Test
1. Why this person need my information?
2. Who this person is?
.png)
.png)
Example of decision making in low-fi stage
.png)
Example of decision making in mid-fi stage
.png)
Example of decision making in hi-fi stage
Comforting: enable users to request information anonomously
Easy: provide default request
.png)
Users receive both an email and a message. This experience is:
Trustworthy: provide that peer's profile keywords to build trust
Convincing: provide professional reasons for the request.
β
.png)
User receives emails and update profile
%20(1).png)
User receives messages on the website and update profile
In this project, I was the product owner and need to prioritize the many design ideas.
β
Challenge Passed
I successfully made my own metrics of prioritization, and my recommendation for project prioritization was recognized by the team.
βοΈ
Its Value to Team
I built a framework of prioritization that can be used in the future.
βοΈ
Its Value to Myself
I figured out my own way to approach and dissect projects. That clarity guides my energy allocation to projects and also helps my design.
Since this was an explorational project, there was no clear expectation of the final deliveries at the beginning. As the owner of this project, I should have set clear expectations of the delivery and success metrics. I failed to do that.
β
π΄
Challenge Failed
I didn't clearly define the scope of this project, didn't set success metrics, and spent much time on designs that were unrelated to the primary goal.
βοΈ
If given another chance
To set clear and feasible project scope, I will:
1. Consider the resources and time given;
2. Only build the MVP and design the others in later iteration with online data;
3. Use the MoSCoW framework to clearly define what to do;
4. Set success metrics.
"Great and Engaging presenter......"
"Best presentation......"
.png)
"Over the summer, Teresa took her UX skills to Another Level......"
"Great at gathering feedback..."
